No, I totally understand. You are correct in that Slony will not replicate the schema for the table, only the data. Generally when bringing up a new node, you put an un-slonified copy of the schema on the subscriber (your raw schema, as if you were running it on one machine), then when you init() the node, slony will slonify the subscriber schema to suit its needs. Then you subscribe the node to the table. Then, you have to let slony make the initial data copy over- once the initial copy is complete, modifications to the master will just start flowing over.
It sounds like you got it to work by a fluke. /kurt On Aug 30, 2007, at 2:38 PM, Benezra, Eric wrote: > I am running Slony version 1.2.10 so I don't have to specify the > listen > paths (even though I did anyway). > > The table I am trying to replicate still didn't work this way. > However, > as soon as I created a blank table of the same name with the same > columns on the slave database, the data replicated over. Is this how > it's supposed to work (create blank tables and let Slony replicate the > info in those tables?) or should it be replicating the whole table > straight to the slave? > > Sorry if that sounded confusing. > > Thanks!! > > Eric > > > > -----Original Message----- > From: Kurt Overberg [mailto:[EMAIL PROTECTED] > Sent: Thursday, August 30, 2007 1:02 PM > To: Benezra, Eric > Cc: [email protected] > Subject: Re: [Slony1-general] Question about replication script > > Okay, thats what I figured. Perhaps when you ran that script against > the slave also, you somehow overwrote the first, correct subscription? > Seems like somehow you may have subscribed a table to itself, which > would, of course, not work so well. I would also check your listen > path > descriptions: > > http://www.slony.info/documentation/listenpaths.html > > and make sure they are properly configured. > > /kurt > > On Aug 30, 2007, at 8:14 AM, Benezra, Eric wrote: > >> Sorry for the confusing wording. What I meant was that, on the >> Master, > >> under subscriptions, the table is there (along with a subscription to >> the slave node). >> >> On the Slave node, there is a subscription also but also to the Slave >> node (to itself?) and no table present. So I was just wondering if it >> had something to do with my replication script or possibly something >> with the communication between the two machines >> >> >> >> >> -----Original Message----- >> From: Kurt Overberg [mailto:[EMAIL PROTECTED] >> Sent: Wednesday, August 29, 2007 8:15 PM >> To: Benezra, Eric >> Subject: Re: [Slony1-general] Question about replication script >> >> Someone please correct me if I'm wrong, but you should only have to >> run this script once, generally against the master (I believe is the >> convention, but I don't think it matters). I'm not sure what you >> mean > >> when you say the table appears on the slave, then you later say that >> the table does not appear on the slave? >> >> Hope this helps... >> >> /kurt >> >> > > Kurt Overberg > [EMAIL PROTECTED] > > > Kurt Overberg [EMAIL PROTECTED] _______________________________________________ Slony1-general mailing list [email protected] http://gborg.postgresql.org/mailman/listinfo/slony1-general
